JP2013127641A - Optical splitter device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an optical splitter device and method.SOLUTION: An optical splitter device includes a waveguide 102 having a wall 104 forming a large hollow core. The waveguide 102 is configured to guide an optical signal via a large hollow core 100. An optical tap 108 is formed so as to penetrate at least one wall 104 of the waveguide 102. Furthermore, a prism 112 is disposed in the large hollow core 100 of the waveguide, and aligned with the optical tap 108. A part of an optical signal is guided via the optical tap 108 to the outside of the waveguide 102 by arranging a splitter coating 110 on the prism 112.

Description

  As the speed of computer chips on circuit boards has become even faster, communication bottlenecks in chip-to-chip communication are becoming a more serious problem. One possible solution to avoid bottlenecks is to use optical fibers to interconnect high speed computer chips. However, most circuit boards contain many layers and the required manufacturing tolerances are often less than 1 micron. Physical placement of the optical fiber and connecting the fiber to the chip is too inaccurate and time consuming to be widely adopted in the circuit board manufacturing process.

  Passing optical signals around and between circuit boards can add significantly to the extra complexity. Thus, the marketable optical interconnection between chips has proven to be a phantom despite being necessary for broadband data transfer.

US Patent Application No. 11 / 832,559

2 is a cross-sectional view of a large core hollow waveguide coupled to a single mode laser and a beam splitter. FIG. It is a figure which shows the walk-off of the beam which arises in a beam splitter. FIG. 6 is a side view of a multimode laser with a collimating lens coupled to a large core hollow waveguide having a reflective interior and a coupling device to the offset large core hollow waveguide. 2 is a side view of a dove prism used as an optical splitter in a waveguide according to one embodiment. FIG. 2 is a side view of a dove prism optical splitter inserted in a waveguide according to one embodiment. FIG. 2 is a side view of an inverted dove prism used as an optical splitter in a waveguide according to one embodiment. FIG. 2 is a side view of a right angle prism used as an optical splitter in a waveguide according to one embodiment. FIG. 2 is a side view of a right angle prism used as an area-based optical splitter in a waveguide according to one embodiment. FIG. 3 is a flowchart illustrating a method for constructing an optical prism splitter in a waveguide according to one embodiment. 1 is a side view of a waveguide patterned and metallized on a silicon substrate according to one embodiment. FIG. 2 is a side view of a silicon substrate bonded to a support substrate using an adhesive layer according to one embodiment. FIG. FIG. 5 is a side view of a waveguide and a slot cut into a substrate according to one embodiment. FIG. 4 is a side view of a prism inserted into a slot in a waveguide on a substrate and a cover mounted on the waveguide according to one embodiment. FIG. 6 is a cross-sectional view of a prism inserted into a slot in a waveguide, sized in a gradually increasing arrangement, according to one embodiment. FIG. 6 is a diagram using prisms with different shapes inserted into slots in a waveguide according to one embodiment. 2 is a cross-sectional view of various sizes of splitter prisms injection molded or stamped into a waveguide wall according to one embodiment. FIG. FIG. 3 is an illustration of an area-based prism splitter in which a prism splitter is formed with a waveguide cover as one embodiment. It is sectional drawing of the area base splitter with which a prism is installed in an optical tap in one Embodiment.

  One method for forming an optical interconnect between computer chips on a circuit board is to use an optical waveguide formed on the circuit board. Since waveguides can be formed on circuit boards using lithography, mechanical, or similar processes, optical waveguides are superior to optical fiber communications that interconnect electronic devices. The waveguide is typically formed on the circuit board by a substantially light transmissive material such as a polymer and / or dielectric. Optical waveguides fabricated using lithography or similar processes can also be formed on other types of substrates that are not mounted on a circuit substrate, such as silicon wafers with microchips. As an additional example, the optical waveguide may be formed on a flexible substrate to form a ribbon cable having one or more optical waveguides. The optical waveguide disclosed herein is formed on a substrate using lithography or a similar process.

  Forming optical waveguides in this manner can provide interconnections built with the physical tolerances necessary for use on modern multilayer circuit boards. However, polymers, dielectrics and other materials that can be used in the manufacture of chips and circuit boards to form waveguides mounted on the substrate are typically significantly more lossy than optical fibers. In fact, the magnitude of the loss in the waveguide mounted on the substrate was one of the factors limiting the acceptance of interconnects by optical waveguides. The polymer used to construct the waveguide can have a loss of 0.1 dB per centimeter. In contrast, optical fiber loss is around 0.1 dB per kilometer. Therefore, the polymer waveguide may have a loss several orders of magnitude greater than the loss of the optical fiber.

  Furthermore, typical waveguides are usually made to have dimensions that are approximately proportional to the wavelength of light they are designed to transmit. For example, a single mode waveguide configured to transmit 1000 nm light may have a maximum dimension of 1000 nm to 5000 nm (1 μm to 5 μm). Connecting waveguides of this size is expensive and difficult. Due to the cost of forming and connecting waveguides, their use in the most common applications has traditionally been low. Multimode waveguides can have dimensions as large as 20-60 μm with respect to the core region. Both single and multimode waveguides have a relatively high numerical aperture (NA). The numerical aperture determines the diffusion of the beam from the emission fiber. Therefore, when NA is larger, as a result, the coupling as a function of separating the fibers becomes insufficient. Using this type of waveguide, it is difficult to achieve splitting and tapping of the guided light beam.

  A substantial improvement over prior optical waveguides formed using polymers or dielectrics is that a large core hollow waveguide configured to guide a coherent light beam 1204 as shown in FIG. 1200 is used. The large core hollow waveguide may have a diameter (or width and / or height) comparable to 50 to 150 times or more of the wavelength of the coherent light beam configured to guide. it can. The large core hollow waveguide can have a cross-sectional shape that is square, rectangular, circular, elliptical, or any other shape configured to guide an optical signal. Furthermore, since the waveguide is hollow, the light beam basically travels in air or in vacuum at the speed of light.

  FIG. 1 shows a laser 1202 that emits a single mode beam 1204 in a waveguide 1206. A splitter 1208 is used to redirect a portion of the light beam, which is represented as a reflected beam 1214 to the orthogonal waveguide 1212. The remaining light beam, represented as transmit beam 1210, can continue in the same direction as the original beam 1204.

  It can be seen in FIG. 2 that a significant amount of beam walk-off can occur within the beam splitter 1208. Beam walk-off is a phenomenon caused by a change in the refractive index between a hollow waveguide (having a refractive index of air or vacuum) and the beam splitter. For example, a beam splitter made of silicon dioxide has a refractive index of 1.45. A beam splitter made of Si3N4 has a refractive index of 2.20.

  The beam is refracted as it enters the beam splitter. The angle at which refraction occurs depends on the difference in refractive index between the waveguide and the beam splitter. Beam walk-off is the distance traveled by the beam due to refraction. This distance is typically proportional to the thickness of the optical device (in this case the beam splitter). Beam walk-off can cause mode displacement in the hollow metal waveguide, resulting in loss due to higher mode excitation near the edge of the waveguide. Lateral displacement can even guide the transmission beam 1210 out of the section 1213 of the hollow metal waveguide that appears after the beam splitter.

式１を２に代入すると、以下の数式が得られる。 An enlarged view of the optical tap is shown schematically in FIG. The incident angle is expressed as θ1 in the air and θ2 in the tap means. The thickness of the optical tap is shown as d, the length of the ray in the tap as s, and the walk-off distance as x. Based on the shape of the beam incident on the tap means,

Substituting equation 1 into 2, the following equation is obtained.

ウォークオフ距離ｘを得るために、

式（４）を使用してかつスネルの法則に従って、

ビームの入射角、タップに対する媒介手段の屈折率、タップの厚さ、およびタップによって生じるウォークオフ距離の間の関係を得ることができる。

To get the walk-off distance x,

Using equation (4) and following Snell's law

A relationship can be obtained between the angle of incidence of the beam, the refractive index of the mediator relative to the tap, the thickness of the tap, and the walk-off distance caused by the tap.

  The beam walk-off can appear exaggerated when using relatively small waveguides. Even when a relatively thin beam splitter 1208, for example about 250 μm (1/4 millimeter) thick, is used in a 50 μm waveguide, the beam walk-off of the transmitted beam traveling through the beam splitter is as much as 75 μm, ie There may be a lateral shift of 1.5 times the thickness of the waveguide. To compensate for the lateral deviation of the beam walk-off, the waveguide can be shifted as shown in FIG.

  Manufacturing can be complicated by shifting the position of the waveguide at each branch point to compensate for beam walk-off. One alternative to minimize beam walk-off is to minimize the beam splitter thickness. This is fully disclosed in co-pending US patent application Ser. No. 11 / 832,559, filed Aug. 1, 2007. However, using extremely thin beam splitters can lead to additional manufacturing complexity. In some embodiments, using hollow metal waveguides that are offset in position can be economical over other methods, such as using thin beam splitters.

  If the path of light through the waveguide is not substantially straight, significant loss can occur in the waveguide. Bends or bends that occur in the waveguide can cause the light beam to bounce off an undesirable number of times, which can cause a significant amount of attenuation. Mirrors, splitters and lenses can be used so that optical signals can be routed in different directions.

  To reduce losses in the hollow core waveguide, a reflective coating 1413 can be applied to cover the interior of the waveguide 1400, as shown in FIG. The reflective coating can be formed using plating, sputtering or similar processes as can be appreciated. If the hollow waveguide has a polymer or other material with a low melting point, the reflective coating can be applied using low temperature processes such as sputtering, electroplating or thermal evaporation.

  The reflective coating 1413 may be composed of one or more layers of metal, dielectric, or other material that is substantially reflective at the wavelength of the coherent light beam. The metal can be selected according to its reflectivity. A highly reflective layer covering the channel is desirable. For example, the reflective layer may be formed using silver, gold, aluminum, or any other metal or alloy that can form a highly reflective layer. Alternatively, the reflective layer may be a dielectric stack that may be formed from one or more layers of dielectric material that are substantially reflective at a selected wavelength. Before the reflective layer is formed, uncoated hollow channels may be subjected to heat reflow to smooth out any roughness of the surface. The reflective layer may also undergo a similar process of smoothing the surface roughness of the reflective layer that may occur during heat reflow or vapor deposition processes. Electropolishing can also be used to smooth the reflective metal surface.

  If the light guide device is not sealed, the reflective coating 1413 may oxidize over time. Oxidation of the reflective coating can significantly reduce its reflectivity. A protective layer 1411 can be formed over the reflective coating to act as a protective cover to mitigate or eliminate the decrease in reflectivity of the metal coating. The protective layer can comprise a material that is substantially transparent at the wavelength of the coherent light beam. For example, the protective layer can be formed of silicon dioxide or any other material that can form a substantially hermetic bond over the reflective coating. The protective layer further reduces the propagation loss by further separating the light beam propagating from the lossy reflective layer.

  A hollow waveguide having a reflecting surface has a different action from a solid waveguide. Hollow waveguides reflect from the reflective layer rather than by total reflection between a higher index core region and a lower index cladding region, as typically occurs in solid waveguides such as optical fibers. It works using the principle of attenuated total reflection guiding light by. The light beam in the hollow waveguide can be reflected at an angle greater than that required for total reflection, as can be appreciated.

  Ideally, a single mode laser is typically used to direct coherent light into the hollow waveguide. However, single mode lasers are relatively expensive. Lower cost multimode lasers, such as vertical cavity surface emitting lasers (VCSEL), can be beneficial for transmitting high data rate signals over relatively short distances using hollow waveguides with a reflective inner surface I know that. To connect between chips and circuit boards, a high data rate signal can be guided into a large core hollow reflective waveguide, for example using a multimode laser. The use of multimode lasers can significantly reduce the cost of optical interconnects and allow a wider variety of electronic devices to be interconnected. However, multimode laser output can be significantly lossy when coupled directly to a hollow metal waveguide due to multiple reflections of higher order modes that propagate at large angles.

  In order to overcome the attenuation of higher order modes emitted from the multimode laser 1402, a collimator 1404 can be placed in the path of the multimode coherent beam 1406 emitted from the laser. The collimator may be a single collimating lens or a series of lenses. In one embodiment, the collimator can be configured as a ball lens. The ball lens can have an anti-reflective coating.

  The collimator 1404 collimates a multi-mode beam or light beam 1406 emitted from the laser 1402 that produces a plurality of various modes, such that the plurality of modes travel substantially parallel through the large core hollow waveguide 1400. A collimated beam 1408 is formed. Use multimode beam collimation to effectively couple a multimode laser into a hollow metal waveguide in a low loss mode by emitting light that is approximately parallel to the waveguide, and the number of reflections that occur in the waveguide Can be substantially reduced. The reflection of the collimated beam that occurs in the waveguide typically occurs at a relatively shallow angle with respect to the waveguide wall, thereby minimizing the number of reflections in the waveguide and thus in the hollow waveguide. The attenuation of the light beam is reduced.

  FIG. 3 shows a system in which a multimode laser 1402 emits a multimode coherent light beam 1406. A multi-mode coherent light beam has a plurality of light rays with increasing angles. As previously mentioned, the light beam is transmitted through a collimator 1404 so that it can be substantially parallel within the large core hollow waveguide 1400. The collimator may be a single lens. Alternatively, the collimator may be composed of a plurality of lenses configured so that the rays of the multimode coherent beam can be made substantially parallel.

  The substantially parallel reflected portion 1414 of the multimode coherent light beam is coupled to the first large core hollow using a coupling device 1410 optically coupled to the first and second waveguides. The direction may be changed from waveguide 1405 to second large core hollow waveguide 1409. The coupling device can be configured to redirect at least a portion of the light beam from the first waveguide to the second waveguide, while the remaining energy is in the first waveguide. Makes it possible to stay. According to one aspect of the present invention, the width and reflectivity of the coupling device can be selected to achieve the desired beam walk-off magnitude.

  1 and 3 present exemplary diagrams for directing a collimated laser beam through a large core hollow waveguide, but the use of a large core hollow waveguide can be The use is not limited to a collimated beam or a coherent light beam. For example, non-parallel laser light can be introduced into a large core hollow waveguide. Higher order modes are inevitably sorted out in the waveguide due to the higher number of reflections that higher order modes experience. Thus, even a non-parallel light beam input to the first waveguide 1405 can be split into a transmit beam 1412 and a reflected beam 1414 as a substantially parallel light beam. It is possible to come out of the tube. Large core hollow waveguides may also be used with substantially coherent light emitted from a light emitting diode or another source of substantially coherent light.

  As mentioned above, the beam walk-off is the magnitude of the beam offset in the waveguide caused by the refraction of the light beam in the coupling device. After the offset transmission portion 1412 of the multimode coherent light beam is shifted in the coupling device 1410, the third large core hollow waveguide 1407 can be shifted to receive this portion. However, fabricating a shifted or offset waveguide as shown in FIG. 3 can complicate the entire manufacturing process and increase the overall manufacturing cost.

  FIG. 4 shows an example of an optical splitter device that can split a light beam without the need for an offset waveguide or a more expensive thin film beam splitter. The exemplary optical splitter apparatus uses a prism to correct the beam walk-off. The optical splitter uses a waveguide 102 having a wall 104 that forms a large hollow core 100. The waveguide is configured to guide the optical signal through a large hollow core. The waveguide walls can be provided with a metal coating 106 for the purpose of reflecting and guiding optical signals, as described above, or an internal reflection coating otherwise.

  The optical signal 130 or beam used by this system may be generated from either a single mode or multimode laser. The laser can use infrared, visible light, or other useful light spectrum. Single mode lasers are typically used in conjunction with hollow waveguides to minimize reflections. However, it has been discovered that a multimode laser can be combined with a collimator to provide a multimode coherent light source having collimated light rays. The use of a coherent multimode laser can significantly reduce manufacturing costs. Furthermore, the use of a coherent multimode light source makes it possible to transmit a high data rate signal using a waveguide. Other forms of substantially coherent light, such as light emitting diodes or infrared light emitting diodes, can also be used.

  A collimated multimode coherent light beam can occupy a significant portion of a large core hollow waveguide. In order to minimize contact between the waveguide and the multimode beam and reduce the number of reflections, the beam may be directed near the center of the waveguide.

  An optical tap 108 can be formed through at least one wall of the waveguide 102. The optical tap can cut open the top of the waveguide so that a portion of the optical signal can be split from the main optical signal and routed through the optical tap. An optical sensor 120 can also be provided to detect the optical signal reflected through the optical tap.

  A prism 112 can be placed in the large hollow core 100 of the waveguide 102 and this prism may be aligned with the optical tap 104. The prism is configured to redirect the optical signal passing through the prism back into the waveguide using a walk-off created by reflecting the transmission beam away from the inner surface of the prism. The transmitted optical signal 134 or beam is in the waveguide in the same orientation as it enters the prism or in another orientation (eg, different angle and / or polarization) defined by the prism and the coating used on the prism. Can be redirected.

  FIG. 4 shows that the dove prism 112 can be the type of prism used to redirect the beam back to the same channel (the channel that is not offset), but other prism shapes and Polygonal prisms can also be used. The prism can be made of an optical material that is transparent to the wavelength of interest. The light wavelength of interest can be, for example, between 1350 nm and 1500 nm, or between 850 nm and 980 nm. Examples of prism materials may be optical silicon glass, optical plastic, quartz and other types of crystals useful for optical purposes.

  The prism can have a splitter coating 110 on the prism to reflect a portion of the optical signal 132 out of the waveguide via the optical tap 108. The splitter coating may be a dielectric coating such as silicon dioxide, tantalum dioxide, titanium dioxide, multilayer dielectric, thin metal coating or any other known splitter coating. The coating on the prism is preferably unaffected by polarization in applications that polarize independent beam splitters. The type and thickness of splitter coating used depends on the desired split ratio and polarization characteristics of the coating.

  In addition to the splitter coating 110 on the light receiving surface, the dove prism can have an anti-reflective coating 116 on the exit surface. This anti-reflection coating allows the transmission beam to effectively pass out of the prism. The dove prism may also be configured to have total reflection on the base 114 of the dove prism. Total reflection can be formed by forming or leaving an air gap 210 between the silicon substrate and the prism. Total reflection can occur when the beam moves from a medium with higher reflectivity to one with lower reflectivity as in the current situation. Alternatively, the base of the dove prism can be metallized to achieve the desired reflection. The incident optical signal or beam that is not split is transmitted into the prism and reflected away from the base 114 of the dove prism. The transmitted beam 134 then passes through the prism, and the walk-off is compensated by symmetrical refraction in the air to the prism medium and the medium to the air interface, so that again after passing through the prism, there is no walk-off again in the waveguide. enter in.

  As mentioned, the ability to redirect the transmission beam back into the waveguide reduces the complexity of the waveguide path that would otherwise inevitably cause a walk-off, so It is beneficial. By providing a beam splitter that returns the beam into the unshifted waveguide, the waveguide path can be simplified. Specifically, by eliminating the need to form offset waveguides or similar adjustment devices, the entire waveguide system can be made easier and overall attenuation can be mitigated.

  FIG. 5 is an example of a method for fabricating a device in which a dove prism is inserted into a waveguide. Initially, a large core hollow waveguide channel is formed. An opening can then be formed in the sidewall of the waveguide. The opening may be formed using a sawing method, laser ablation, etching or photolithography process. The dove prism 112 can then be inserted into the waveguide channel and secured in place. A waveguide cover with slot openings can then be added to the waveguide channel so that the reflected optical signal can be routed out towards the optical sensor or detector 120.

  In the displayed embodiment of FIGS. 4 and 5, the optical sensor 120 is shown substantially offset from the position of the waveguide. However, the optical sensor may be placed directly adjacent to the waveguide.

  FIG. 6 illustrates the use of an inverted dove prism 310 in the waveguide. Although an inverted dove prism can function using a walk-off, different surfaces of the dove prism may be used to split the optical signal. The inverted dove prism can have a splitter coating 112 on the inverted base of the inverted dove prism. Anti-reflection coatings can be placed on both the entrance surface 110 and the exit surface 116 of the inverted dove prism. This configuration allows the incident optical signal or beam to be split into a reflected beam 132 that reaches the sensor 120 and a transmission beam 134 that is returned into the waveguide.

  As shown in FIG. 7, other types of prisms can be used to achieve a similar result of modifying the beam walk-off. For example, the prism disposed in the waveguide may be a triangular prism. More specifically, the triangular prism may be a right-angle prism 410. The right angle prism can have a splitter coating 412 on the entrance surface and an anti-reflective coating on the exit surface 416. The incident optical signal or beam is split into a reflected beam 432 that reaches the sensor 420 and a transmit or return beam 434 that is returned into the waveguide.

  The hypotenuse 414 of the right-angle prism can use total internal reflection to redirect the beam refracted back into the waveguide by walk-off. After the beam is redirected by the first magnitude walk-off, the beam bounces off the hypotenuse at the angle of incidence, and then the second magnitude walk-off returns the transmitted beam back into the same waveguide channel. Turn to. The reflected beam may be transmitted back into the waveguide at the same angle as it enters the prism. Alternatively, the light beam may travel out of the prism at an angle that is different from the angle at which it originally entered the prism, depending on the actual geometry of the prism and the prism coating.

  FIG. 8 shows the use of a triangular or right-angle prism as an area-based beam splitter. The guided beam fills the waveguide as it propagates, so if a particular area of the waveguide is formed as a 45 degree angle reflective surface to split the optical signal by 90 degrees, one of the beams The part can be reflected. The split ratio roughly depends on the ratio of the propagation region to the reflection region in the waveguide that depends on the mode shape of the propagating beam.

  To construct a waveguide structure with a plurality of split ports with varying split ratios, the plurality of triangular prisms 510, 518 can have a metallized reflective layer applied to the hypotenuse 512 of the prism. . These prisms can be inserted into the waveguide 100 at gradually increasing heights. Insert the prism into the waveguide using several cross slots that can be sawed or machined into the waveguide and silicon substrate to form steps 550, 558 on which the prism is placed can do. The prisms may be fixed to the silicon substrate as they are inserted into the step, or prisms of varying height or size may be formed directly on the substrate.

  The area prism is inserted into the waveguide sufficiently far away to reflect the desired portion of the light beam towards the detector 120. For example, the first area prism can reflect the lowermost portion 530 of the light beam. If the prisms are then sufficiently close together, the second area prism can reflect the next height portion 536 of the light beam. There may be additional prism splitters or reflectors not shown in the drawing that reflect the remainder of the light beam 540 until all the light beam is reflected out of the waveguide. The approximate division ratio is determined by the region of the prism that covers a part of the waveguide cross section.

  If the first and second prisms are sufficiently separated, the remaining portion of the beam that is not reflected by the area-based splitter completely fills the waveguide channel during propagation, so that the percentage of the area occupied by the second prism Determines the split ratio. The presented design can be particularly useful in optical bus configurations where a series of beam splitters are incorporated into the waveguide to route the optical signal.

  An exemplary method for fabricating an optical splitter device for a waveguide having a large hollow core will now be described, with each process described in relation to a cross-sectional fabrication drawing in separate figures. FIG. 9 is a flowchart showing an initial process of forming a hollow channel in the silicon substrate as block 610. FIG. 10 shows the formation of a waveguide 710 on top of a silicon substrate 720, such as a silicon wafer. The waveguide can be formed by sawing, stamping, laser patterning, photolithography or other semiconductor manufacturing techniques. For example, a waveguide can be formed by patterning a photoresist on a silicon substrate and then using an exposure process to remove unwanted portions of the silicon substrate. As a result of this step, a hollow channel with a bottom and side walls can be formed. For example, patterning a silicon substrate can further include applying a photoresist, utilizing a photolithography process, utilizing a dry etching process, and using a cleaning process.

  Alternatively, the channel can be sawed into the silicon substrate by a sawing process. For example, a dicing saw can be used to cut the channel of the waveguide. Dicing saws utilize high speed spindles adapted to thin diamond blades or diamond wires to dice, cut or form grooves in semiconductor wafers, silicon, glass, ceramics, crystals and many other types of materials You can do it.

  Once the waveguide channel is formed, metal can then be deposited in the hollow channel to form a hollow channel reflection as in block 620 (FIG. 6). Metal deposition can include applying an AIN (aluminum nitride) protective layer and then applying reflective silver with titanium as a buffer layer to increase the adhesion of the metal layer of the substrate. . As described above, other metals can also be used to form the reflective layer. This reflective surface in the channel forms the waveguide channel.

  FIG. 11 illustrates optional steps of the method including adhesion of another silicon wafer 800 to the first silicon substrate or the first silicon wafer. The second wafer can be bonded by the adhesive layer 810. The reason why the second wafer can be bonded to the first wafer is to increase the overall substrate depth in preparation for the next sawing step. Furthermore, the adhesion between the two wafers tends to relieve stress associated with deep cutting and provides mechanical integrity. If a sufficiently thick silicon substrate is used, this process may not be required.

  Another process is to saw the depth slot across the waveguide as in block 630 of FIG. FIG. 12 shows that the depth of each depth slot 910, 912, 914 gradually decreases as the depth slot distance increases from the optical signal source. The depth slot can be sawed approximately perpendicular to the waveguide, or can be sawed at other angles when a particular optical application is selected. The slot depth depends on the prism size, the area ratio required to achieve the desired split ratio for a given prism, and the mode characteristics of the guided beam in the waveguide channel. .

  A right angle prism can then be inserted into the depth slot as in block 640 of FIG. As shown in FIG. 13, each prism 950, 952, 954 can have a metal coating or other highly reflective coating on the hypotenuse of the prism. The right angle prism is thus configured such that the hypotenuse is directed to the incident optical signal. These area-based splitters can then reflect the optical signal or beam out of the waveguide based on the area of the waveguide that each splitter occupies. The configuration depicted in FIG. 13 is not necessarily drawn to scale, but with a stepped support structure and a selected section of the waveguide from a selected point outside the waveguide (eg, a sensor). It is merely an example of an area-based splitter structure using a prism with a hypotenuse that reflects almost all light.

  The final process is the step of applying a cover over the hollow channel as in block 650 of FIG. Although the cover can be applied before the sawing is performed, the order of the processes described in this disclosure can be performed in any effective order. FIG. 13 further shows a waveguide cover 960. The waveguide cover includes a slot that can direct light from the right-angle prism to a point outside the waveguide. For example, a point outside the waveguide may have an optical sensor, a microlens array or other waveguide channel. An alternative way to achieve a variable split ratio is to use prisms of different sizes with a fixed depth slot. In this method, the slot for inserting the prism is cut at a certain depth, while the inserted prism is formed to gradually increase in height to provide the desired split ratio.

  FIG. 14 is a cross-sectional view of a waveguide 1410 formed with a substrate 1412 having a cover 1420. According to one embodiment, prisms 1414, 1416, 1418 of increasing size are inserted into the slots. In this type of area-based prism division, the size of the prism can be increased instead of increasing the height of the steps formed in the substrate as in other embodiments. Therefore, a uniform step height can be used within the substrate, and the prism size can be increased as the prism is placed further away from the light beam source to change the split ratio. Typically, these types of area splitters are placed relatively close together to avoid expansion of the optical signal in the waveguide between each splitter.

  FIG. 15 is a cross-sectional view of prisms with different shapes inserted into slots in a waveguide according to one embodiment. Use many prism shapes with surfaces suitable for reflecting the beam out of the waveguide at the desired angle, such as right angle prism 1502, dove prism, equilateral triangle 1506, hexagon, octagon and other polygons 1504 can do.

  FIG. 16 is a cross-sectional view illustrating the use of various sized prisms 1610 used in a waveguide 1620 that can be fabricated on a substrate using injection molded plastic or similar materials. The surface of the prism can be metallized as described above. Alternatively, such a prism can be stamped or sawed from the waveguide wall to form a splitter prism. The area-based splitter described above can be made directly on or in the substrate in an integral form using injection molding, stamping, sawing and similar manufacturing processes.

  FIG. 17 shows an area-based prism splitter formed with the cover or wall of the waveguide 1710. These reflective prism splitters 1710, 1720, 1730 can be formed in the waveguide walls using stamping, mechanical stamping or similar manufacturing methods.

  FIG. 18 shows a cross-sectional view of an area-based splitter in waveguide 1800 where the prism is installed via an optical tap. The waveguide cover 1810 can have prism mounts 1820, 1830 that are installed within the optical tap and that support a triangular or polygonal prism within the optical tap region. The hypotenuse 1840 of the triangle can form a reflection using metallization or other techniques. Thus, the incident beam enters the prism through the surface with the anti-reflective coating and is reflected out of the light tap using the reflective hypotenuse.

  While the method described above is effective for area splitters such as right angle splitters, it is also possible to use similar sawing methods or photoetching steps to form the location for the dove prism. The formation of the waveguide and the metallization process used to incorporate the dove prism into the waveguide are substantially the same. However, the slots sawn into the waveguide for the dove prism each have the same depth, after which the dove prism is inserted into the slot of the waveguide. The dove prism is configured to transmit an optical signal or a portion of light, so that the dove prism can be inserted directly into the desired location within the waveguide.

  An alternative embodiment can use a dove prism as an area splitter. This can be done by changing the size of the dove splitter in the same way as has been shown for triangular prisms. In this way, the dove splitter can be configured so that a selected amount of light can pass past the dove prism and reach another dove splitter.

  Large core hollow waveguides can be used to interconnect electronic devices located on one or more circuit boards or silicon wafers. The electronic device may have an electrical output and an input that is converted to an optical output for transmission through the optical waveguide. Alternatively, the electronic device may be an optical device that transmits and receives optical signals that do not need to be converted. Large core hollow waveguides with a reflective coating inside the waveguide can greatly reduce the loss of optical signals induced through the waveguide compared to solid core waveguides. . The reflective coating inside the hollow waveguide can minimize losses caused by reflection of the optical signal in the waveguide. The use of a prism splitter corrects the walk-off and allows the beam transmitted through the prism to return to the same waveguide at the same angle that this beam is received. In addition, the prism can be provided with a reflective coating to direct a portion of the optical signal based on the area occupied by the prism in the waveguide.

  The foregoing examples are illustrative of the principles of the present invention in one or more particular applications, but do not embrace the inventive capabilities and do not depart from the principles and concepts of the present invention. It will be apparent to those skilled in the art that many modifications can be made to the details. Accordingly, the invention is not intended to be limited except as by the claims set forth below.

  The embodiments of the present invention are listed below.

1. A waveguide (102) having a wall (104) forming a large hollow core (100) and configured to direct an optical signal (130) through the large hollow core;
An optical tap (108) formed through one wall of the waveguide;
A prism (112) in the large hollow core of the waveguide and aligned with the optical tap;
A splitter coating (110) on the prism for guiding a portion of the optical signal out of the waveguide through the optical tap;
An optical splitter device comprising:

2. The prisms are dove prisms or inverted dove prisms, each configured to redirect an optical signal transmitted through the prism back into the waveguide by walk-off.
2. The optical splitter device according to item 1 above.

3. The dove prism has a splitter coating on the light-receiving surface, an anti-reflection coating on the exit surface, and total reflection at the base of the dove prism;
3. The optical splitter device according to item 2 above.

4). The inverted dove prism has a splitter coating on the inverted base of the inverted dove prism, and has an antireflection coating on the entrance and exit surfaces of the inverted dove prism,
3. The optical splitter device according to item 2 above.

5. The prism disposed in the waveguide is a triangular prism;
2. The optical splitter device according to item 1 above.

6). The triangular prism has a reflective metal-coated coating applied to the hypotenuse of the triangular prism, and the triangular prism returns an optical signal transmitted through the prism into the waveguide by walk-off. Configured to redirect
6. The optical splitter device according to item 5 above.

7). The prism is installed in the optical tap using a support formed from a waveguide cover,
2. The optical splitter device according to item 1 above.

8). A plurality of area-based prisms that increase in size as the distance from the optical signal source increases;
8. The optical splitter device according to item 7 above.

9. The prism further comprises a plurality of right angle prisms having a reflective metallized coating applied to the hypotenuse, wherein the right angle prisms are used as area base splitters in the waveguide;
2. The optical splitter device according to item 1 above.

10. A plurality of prisms are formed against one wall of the waveguide using a process selected from the group consisting of injection molding, stamping and stamping.
2. The optical splitter device according to item 1 above.

11. A method for fabricating an optical splitter device for a waveguide having a large hollow core, comprising:
Forming a hollow channel in the silicon substrate (610);
Depositing a metal on the hollow channel to make the hollow channel reflective (620);
Sawing (630) a depth slot across the waveguide such that the depth of each of the depth slots gradually decreases as the depth slot distance increases from the optical signal source;
Inserting a right angle prism having a metallized coating on the hypotenuse, wherein the hypotenuse of the right angle prism is oriented toward the incident optical signal into the depth slot (640);
Applying a cover over the hollow channel (650);
A method comprising the steps of:

12 Forming the hollow channel includes sawing the hollow channel into the silicon substrate using a dicing saw;
12. The method according to item 11 above, wherein

13. Adhering a second silicon wafer to the bottom of the first silicon wafer;
12. The method according to item 11 above, wherein

14 The step of patterning the silicon substrate further includes applying a photoresist, utilizing a photolithography process, utilizing a dry etching process, and using a cleaning process.
12. The method according to item 11 above, wherein

15. A method for fabricating an optical splitter device for a waveguide having a large hollow core, comprising:
Forming a hollow channel in the silicon substrate;
Depositing metal on the hollow channel to make the hollow channel reflective and form a waveguide;
Sawing a depth slot across the waveguide, the depth slots each having the same depth;
Inserting a dove prism into the depth slot;
Applying a cover over the hollow channel;
A method comprising the steps of:

Claims (10)

A waveguide having walls forming a hollow core, the waveguide configured to guide an optical signal through the hollow core;
A plurality of prisms spaced apart from each other in the direction in which the optical signal is guided, and protrudes into the hollow core so that the degree of protrusion increases as the distance from the optical signal source increases. , With multiple prisms,
A plurality of optical taps formed through the first wall of the waveguide, each of the plurality of optical taps being aligned with each of the plurality of prisms; A light tap,
And a reflective coating on each of the plurality of prisms for directing a portion of the optical signal to the outside of the waveguide through each of the plurality of optical taps.   The each of the plurality of prisms is mounted in each of a plurality of depth slots formed in a second wall opposite the first wall of the waveguide. Optical splitter device.   The optical splitter apparatus according to claim 2, wherein each of the plurality of depth slots has a depth that decreases as the distance of the depth slot from the optical signal source increases.   3. The plurality of depth slots has a uniform depth, and the prisms increase in size as the plurality of prisms are positioned further from the optical signal source. An optical splitter device according to claim 1.   2. The optical splitter device according to claim 1, wherein the plurality of prisms are formed integrally with a second wall facing the first wall of the waveguide.   6. The optical splitter apparatus of claim 5, wherein the plurality of prisms are formed in the second wall using a process selected from the group consisting of injection molding, stamping, and stamping.   2. The optical splitter device according to claim 1, wherein the first wall is formed by a waveguide cover, and the plurality of prisms are formed together with the waveguide cover.   The first wall is formed by a waveguide cover, and each of the plurality of prisms is mounted in each of the plurality of optical taps using a support formed from the waveguide cover. The optical splitter device according to claim 1, comprising:   The optical splitter device according to any one of claims 1 to 8, wherein the reflective coating is a reflective metal coating provided on an optical signal receiving slope of each of the plurality of prisms.   The waveguide is configured to direct a coherent light beam, and the hollow core has a diameter or width and / or height that is 50 to 150 times or greater than the wavelength of the coherent light beam; The optical splitter device according to claim 1, comprising:

Images